Fresh thinking about telecommunications
View this email in your browser

The Best of 2013

Here is a summary of the key Future of Communications content from 2013.

It’s not (just) about bandwidth

The basic nature of packet networking is to cause and schedule contention. That means capacity is not king: schedulability is (ultimately) supreme. Network operators vary in their ability at scheduling, so bandwidth is not created equal. I elaborated on this in my ITU Future of Broadband workshop presentation. The investment in fibre (and 4G), without attention to the scheduling issues, is delivering us an information super mud track.

Emergence of network (performance) science

The reason we’re in a pickle is that we have yet to achieve the foundation of ‘network science’. There is a widespread belief you can always cheaply over-provision away your scheduling problems, yet there is no “Moore’s Law” for networks. It is dangerous to extrapolate the past, as it means you fall for the ‘frequentist fallacy’.

Indeed, the Internet’s conceptual foundations are fractured. We are entering a qualitatively different era of distributed computing; as such traditional computer networking is dead. We should be looking to new technology paths, and anticipating the rise of the ‘sensible network’ to better support society’s needs. Common objections to offering fit-for-purpose broadband services fail to grasp the paradox of purpose and how to deal with it.

So what is this new ‘network science’? The essence is to understand the role of contention in the third epoch of networking. We can separate out dynamic flow issues from structural ones by grasping the fundamentals of network performance engineering. It is possible to formally and robustly model network performance and economics using ‘The Ladder’ to tie the money and scheduling together.

Network technology

This network science approach is vital, as networks are increasingly like distributed supercomputers, where virtualisation is hard. A failure to grasp its fundamental precepts have sabotaged past efforts at access link bonding, as well as attempts to find viable ways to produce and sell QoS.

The whole TCP/IP approach to networking is just an escaped 1970s lab experiment, and there are fresh paths to follow in network architecture research.

Future of voice

Voice remains an ever-popular theme. We’re seeing the ongoing decline of the old telephony approach to voice, heralding the silence of the landlines. As the epochs turn, we can say goodbye to retrovoice and hello to neovoice. We are leaving familiar territory for a new world where voice is computer-processed data, as described in The Hypervoice Manifesto.

This transition of voice from telephony to new forms provokes some interesting questions, like: Who owns your phone number? How to close the conversation gap between enterprises and their customers?

Wholesale telecoms

My big surprise in 2013 was a sudden uptick in business from the wholesale telecoms sector, one that continues into 2014. I wrote about the opportunity to re-invent wholesale voice and data, which in turn became a conference keynote on fit-for-purpose wholesale data. Meanwhile, the wholesale telecoms business carried on working on IPX, which may not offer the hoped-for salvation.

Broadband policy & economics

We’re in the midst of a socio-economic revolution brought on by the cloud and cheap information everywhere. Given these radical changes, it is time to re-examine broadband regulation. After all, it seems that telcos are becoming slum landlords, extracting rent from their tenants through inefficient and ineffective data caps. Policy-makers struggle to grasp the economics of bandwidth, and cannot make up their minds if network neutrality is nasty or nice.

The future telco

So, when I look into my telecoms crystal ball, what do I see? There are two over-riding issues. The first is to meet the OTT challenge by exploiting the network ‘trading space’ to create fit-for-purpose services. The second is to collapse complexity (and cost) by ‘industrialising’ telecoms operations.

Thanks to all of you for being part of this journey of discovery in 2013. I have plenty of new content ready on the launch pad for 2014, and look forward to sharing it with you soon.

Martin Geddes
PS - You can also view the â€œBest of 2012” from last year, if you want to learn more.

Upcoming speaking engagements

28th-31st January, Miami IT Expo.
24th-27th February, Barcelona - Mobile World Congress.
4th-5th March, London - Connected Business.
1st-2nd April, LondonWebRTC Global Summit.
29th April-1st May, Barcelona - IMS World Forum.
10th-11th June, MunichNext Generation Services Platforms.

Subscription and back issues

This newsletter is free. If it was forwarded to you, and you are not yet subscribed, then sign up hereThe hyperlink to this article and recent newsletter back issues is here.

Copyright © 2013 Martin Geddes Consulting Ltd, All rights reserved.
unsubscribe from this list    update subscription preferences